467 



in the majority of the arctic flowers in my material. Whether 

 this peculiar phenomenon is caused by a natural tension in 

 the stigma or by the agency of an insect I cannot decide, 

 though its frequency makes the first mentioned explanation 

 probable. In the cases mentioned here the stigma has been 



Fig. 10. Finguicula vulgaris, 

 A—C, A flower, seen from different sides; it had two rudimen- 

 tary anthers, but the pistil was normal. D—F, A flower with 

 only one anther; the pistil was normal; D, Side-view; K, front- 

 view; F. back-\aew. G, A flower with a spur, slightly divided 

 into two parts. H, A hair from the inner side of the corolla. 

 A~F. From Tromsø; G— H, From Godhavn in Greenland (E. W.) 



thickly covered with germinating pollen (Fig. 9, B, C, D). Ac- 

 cording to all investigators only very few insects visit the 

 flowers of B. vulgaris, and in the arctic countries, where 

 larger insects are fewer in number than in other regions, the 

 necessity of self-pollination must be exceedingly great. This 

 agrees very well with the usual explanation of the biological 

 circumstances in arctic countries: that self-pollination is, as 

 first mentioned by Warming, very common here. 



